To estimate the intestinal contribution to the first pass metabolism of orally administered midazolam. The hypothesis is that midazolam is metabolized by intestinal cytochrome P450 3 A (P450 3A) enzymes to a clinical significant amount. To measure the effect of P450 3A inhibition by orally administered ketoconazole on the pharmacokinetic parameters of orally and intravenously administered midazolam. The hypothesis is that oral ketoconazole affects the disposition of the oral dose more than that of the intravenous midazolam. Preliminary analysis has shown some racial differences in the metabolism of midazolam. Continued analysis will determine to what extent the significance.